CPSC 433 - Artificial Intelligence

January 16, 2007

Organization

• We created groups and did a rough introduction to the course.

January 18, 2007

Intelligence

• We covered the Assignment requirements in more depth
• We discussed Eliza and its implications on intelligence, check out some instantiations of Eliza and its descendants.

January 23, 2007

Puzzle Definitions

• We talked about the puzzle example from Professor Denzinger's Two Chapters, and defined the problem in terms of the formal definitions to create the first search model.

January 25, 2007

Puzzle Examples

• We continued the puzzle example, working through the first and second search controls for the first search model.

January 30, 2007

Puzzle Examples Again

• We finished the first search model for the puzzle problem. We also gave the formal definitions for the second search model. I will attempt to post the slides I used to illustrate the search protocols in the Handout section shortly.

February 1, 2007

Yet another lab for puzzle examples

• We finished the puzzle search models. We also started the 0-1 knapsack problem, providing a definition of the problem and a general definition of a Genetic Algorithm, finally we looked at the 0-1 knapsack problem as a problem to be solved with a GA.

February 6, 2007

0-1 Knapsack - GA

• We looked briefly at why (and how) genetic algorithms work. We then applied a genetic algorithm search to the 0-1 knapsack problem.

February 8, 2007

0-1 Knapsack - Branch & Bound

• We looked at another solution to the 0-1 knapsack problem, this time a tree based approach.
• Bound is calculated by summing the profit of selected items (and here we can take parts of items) up to a weight(cost) of K. This can result in taking, for example, all of item 1 and 1/2 of item 2. This does not mean that we will take 1/2 of item 2 in our final solution, but it allows us to estimate the best possible result in the subtree below the node we are looking at.

February 13, 2007

Travelling Salesman - Branch & Bound

• We designed a search model to solve the Travelling Salesman problem. I also jumped in and gave a second model that didn't necessarily make things clearer.

February 15, 2007

Travelling Salesman - Branch & Bound

• We looked at our branch & bound paradigm and found that it needed some repair.

February 27, 2007

Travelling Salesman - Branch & Bound - one more time

• We covered some questions relating to the papers and we finished the Branch & Bound problem.

March 1, 2007

Blocks World

• We looked at or-trees to solve problems in Blocks World. Keep in mind that with the example we did the solution to the problem (the actions to go from the start situation to the goal situation) is described in the path (each branch represents a choice of action) from the root node to the node we changed to 'yes'.

March 6, 2007

Midterm Review

• We covered questions relating to the midterm. I recommend you look at the preparation questions from Prof. Denzinger.

March 15, 2007

First Order Logic

• We finished covering the midterm, and started FOL (before I lost the ability to write on the board).

March 20, 2007

First Order Logic Again

• We finish (correctly) Modus Ponens, conversion to CNF and do a larger example.

March 22, 2007

Witches! or maybe Prolog

• We will looked at Prolog as an example of FOL and learn why witches burn.

March 27, 2007

Why it snows

• We used Emycin to look at the things people do to cause snow.

March 29, 2007

• We looked at XML; DTDs and how to fix a broken document.

April 3, 2007

• We looked at CD diagrams.

April 5, 2007

• We will looked at perceptrons as an example of neural nets. The important thing to know is how to do the learning update.